1
\$\begingroup\$

I am using this pre-made PCB in order to bring my 12V power supply to 6V. On the datasheet, it is claimed that the efficiency is about 90% (hence I expected low heat dissipation). With great surprise when I plugged everything together, I've noticed that the PCB heats up quite a bit. I haven't much experience with these things and therefore I would like to know if you'd expect it to heat a lot or not. I am scared that something is wrong.

\$\endgroup\$
8
  • 2
    \$\begingroup\$ How much current are you drawing? How hot? Can you hold your finger on it indefinitely? Does it boil water? \$\endgroup\$ Sep 22, 2015 at 16:23
  • \$\begingroup\$ Right, i forgot to add: \$\endgroup\$
    – Worldsheep
    Sep 22, 2015 at 16:25
  • 1
    \$\begingroup\$ Right, i forgot to add: with no load I have 20-30 mA and about the heat... After 3-4 minutes, if I put my finger on it for 4-5 seconds I start to feel it a bit painful, but still possible to hold my finger on it :). \$\endgroup\$
    – Worldsheep
    Sep 22, 2015 at 16:30
  • 1
    \$\begingroup\$ If it's not shutting itself down, then I wouldn't worry about it. According to the datasheet it has thermal overload protection. \$\endgroup\$
    – DerStrom8
    Sep 22, 2015 at 16:37
  • \$\begingroup\$ But how can it heat so much with an efficiency of over 90%? Is it because I have no load? Do you suggest adding a small heat sink? \$\endgroup\$
    – Worldsheep
    Sep 22, 2015 at 16:39

1 Answer 1

2
\$\begingroup\$

It states on the front page that efficiency is 93%. It also states that you can use it to supply 3A. That means an 18 watt load if you have the output set for 6 volts. A 7% "inefficiency" is about 1.3 watts and given its small size it will get quite warm.

If you have a genuine 30mA flowing into it from the incoming 12V supply then the actual no-load power is only 0.36 watts so, my advice is to expect it to get a bit warmer under full load conditions and make sure you can provide natural convection to remove heat (see graph at top of page 5 of data sheet).

\$\endgroup\$
7
  • \$\begingroup\$ I don't understand what the derating graph is telling... How can i properly choose an heat sink? \$\endgroup\$
    – Worldsheep
    Sep 23, 2015 at 6:06
  • \$\begingroup\$ @Worldsheep I wouldn't be happy with what they say in the data sheet. There are words towards the end of the document that don't really tie in with the graph on page 5. Nothing direct is mentioned about heatsinking. It's nice to find a to220 type regulator that replaces a LM78xx linear type but I suspect you might fair better with a different device with an external inductor. I believe Rohm make some that look useful. \$\endgroup\$
    – Andy aka
    Sep 23, 2015 at 7:36
  • \$\begingroup\$ I first tried to build by own DC/DC converter using a buck step down converter and it didn't heat up as much as this one. The problem was that I am testing on a breadbord and it was very unstable. Finally I had to give up and buy a pre-made module... Wouldn't a device with external inductor cause a similar instability issue? I'll have a look at Rohm! But in your opinion, are there solutions that heat less? \$\endgroup\$
    – Worldsheep
    Sep 23, 2015 at 7:56
  • \$\begingroup\$ There are plenty of solution that heat less. Linear tech and TI spring to mind. I would encourage you to design a small PCB to overcome your breadboard issues. You can always post the PCB design here on SE and get opinions about if there are any potential layout problems. \$\endgroup\$
    – Andy aka
    Sep 23, 2015 at 8:02
  • \$\begingroup\$ The main issue is not the layout which is provided with the datasheet, but the time needed to design it and then produce it just to test it :). That's why I went for a pre-made :/ \$\endgroup\$
    – Worldsheep
    Sep 23, 2015 at 8:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.